Hi Martin, On Mon, Jan 13, 2014 at 07:39:05AM +0100, Martin Pitt wrote: > Hello, > > Matthias Klose [2014-01-13 3:02 +0100]: > > This test is failing this is was introduced in 2012.10.24-6. However the > > log > > doesn't tell what exactly is failing. > > > > - afaics the test is at least missing a dependency on aspcud > > - sending all output to /dev/null is a bad idea for an untested > > test script. > > Attached patch at least fixes those, and the wrong $indir directory > (autopkgtests are run from the source tree root). Please don't do > "2>/dev/null". If your test always writes expected stuff to stderr, > then add "Restrictions: allow-stderr" to debian/tests/control. > Otherwise, if the test is not supposed to write to stderr, don't > redirect it (preferrable to detect unexpected > regressions/warnings/etc.) > > But now further errors appear: > > | $ TMPDIR=/tmp/x sh -x debian/tests/upstream > | + set -e > | + indir=debian/tests/cudf > | + outdir=/tmp/x > | + basename debian/tests/cudf/feature.cudf > | + cudf=feature.cudf > | + outfile=/tmp/x/feature.sol > | + checkfile=/tmp/x/feature.check > | + aspcud debian/tests/cudf/feature.cudf /tmp/x/feature.sol trendy > | *** ERROR: (clasp): Read Error: Line 0, Bad input stream!
thanks for your patch. However, I cannot reproduce that Bad Input Stream error on my machine. What are the versions of the dependencies of aspcud that you had installed when running this test ? Cheers -Ralf. -- To UNSUBSCRIBE, email to [email protected] with a subject of "unsubscribe". Trouble? Contact [email protected]
